Soot generation is an important problem in high-temperature biomass gasification, which results in both air pollution and the contamination of gasification equipment. Due to the complex nature of biomass materials and the soot formation process, it is still a challenge to fully understand and describe the mechanisms of tar evolution and soot generation at the reactor scale. This knowledge gap thus motivates the development of a comprehensive computational fluid dynamics (CFD) soot formation algorithm for biomass gasification, where the soot precursor is modeled using a component-based pyrolysis framework to distinguish cellulose, hemicellulose and lignin. The model is first validated with pyrolysis experiments from different research groups, after which the soot generation during biomass steam gasification in a drop-tube furnace is studied under different operating temperatures (900–1200 °C) and steam/biomass ratios. Compared with the predictions based on a detailed tar conversion model, the current algorithm captures the soot generation more reasonably although a simplified tar model is used. Besides, the influence of biomass lignin content and the impact of tar and soot consumptions on the soot yield is quantitatively studied. Moreover, the impact of surface growth on soot formation is also discussed. The current work demonstrates the feasibility of the coupled multiphase flow algorithm in the prediction of soot formation during biomass gasification with strong heat/mass transfer effects. In conclusion, the model is thus a useful tool for the analysis and optimization of industrial-scaled biomass gasification. © 2021 The Author(s)